118 ALASKAN GLACIER STUDIES
due to a predominance of green schist, while beyond this came a purplish moraine, due to the presence of a purple gneiss in sufficient quantities to mask the scattered red, orange and black fragments in it. There were other shorter, less distinct moraines with curvature parallel to the neighboring more pronounced ridges. The purple moraine, which had a width of about an eighth of a mile, descended steeply to the interior flat, but on this descent were two minor, not very distinct ridges, one red in color, the other white, owing to the presence of many fragments of white granite.
Because of our interest in the peculiar banded moraines we traversed this part of the ice bulb and nowhere found the least evidence of motion. It was a greatly roughened surface, strewn with large angular blocks, but not a single crevasse was found. Here and there ice was seen, especially on the steeper ridge slopes and in the valleys, and everywhere there was clear evidence that ice existed just beneath the moraine surface, and we could usually reach it by thrusting our ice axes into the moraine. The surface was constantly shifting by the imdermining caused by melting of the ice beneath, and probably this is the main reason why there was so little vegetation. Scattered willows, none over two years old, were growing here and there on all the moraine ridges excepting the orange moraine. Many dead plants were seen, and many others with their roots partly exposed by recent undennining. Evidently in such a shifting soil the growth of but' a year or two was all that a plant could attain before the slumpingof the surface removed the soil in which it was growing.
The Interior Flat in 1905. The interior flat had the form of a crescent with the concave side toward the mountain valley, where it was bordered by the steeply-rising red and purple moraine whose crest was about 250 feet above it; on the outer side it was bordered by the concave ridge of the broad, black hornblende gneiss area of the outer bulb, which rose over 200 feet above it. The flat was, therefore, a very notable feature, (PL LI, A) interrupting the succession of crescentic moraines of the inner and outer bulbs. Its length was more than a mile and its width in its widest part fully three quarters of a mile, but narrowing rapidly toward either end. Its elevation was about 100 feet above sea level. The surface of the interior flat was mainly composed of gravels brought by streams issuing from the inner bulb of the Variegated Glacier, but some low ridges and hummocks rose above the alluvial flats. There were areas of depression occupied by lakes, some pits formed by recent settling, and numerous cracks and faults, proving that the settling was still in progress. These facts prove the presence of ice beneath the interior flat, but that it was not in motion was clearly indicated by the general levelness of the floor of the interior flat. Probably the ice beneath the flat was very thin.
It is believed that the interior flat represents the site of a clear ice area, such as is observed in other glaciers of this region, which, when the glacier became stagnant, was so lowered by ablation that it formed a depression in the glacier bulb in which sedimentation ultimately began, thus burying the ice. In 1905 two streams were contributing to the alluvial deposits, one from the northern side bringing the drainage of the northern margin of the Variegated Glacier, the other, and much the larger, emerging (PI. LVH, A) from an ice gorge in the southern portion of the inner bulb, then descending in a fall into a granite gorge, from which it emerged first across moraine, then upon the interior flat. Both streams flowed in branching course across the interior flat, and then, after uniting, passed through a moraine-walled ice gorge in the outer bulb of the glaciers, and finally, uniting with a stream from Orange Glacier, flowing in many distributaries to the fiord over a very large